Management of Brain Parameters
Indications for TTM Based on Prediction of Neurologic Outcome on Admission
Recent guidelines primarily address the termination of resuscitative efforts during performance of TTM.8,9
Although the International Liaison Committee on Resuscitation task force has suggested the use of a combination of tests for this purpose, combined tests predicting outcomes were not evaluated in those guidelines. And in Japan, it is hard to withdraw intensive care from patients who have suffered cardiac arrest because of their social background.
Multiple modalities of testing to predict outcomes are acceptable if they can be applied quickly on admission. In this section, we will review the recommended predictors of neurologic outcomes following OHCA prior to initiating TTM.
Single Test on Admission
The Glasgow Coma Scale (GCS) is very simple, speedy test that can be applied on admission. The initial GCS motor score is an independent predictor of good neurologic outcome at 90 days in patients with OHCA treated with therapeutic hypothermia.10
As a single test, even patients with a GCS motor score of 1 had a 52.2% probability of good neurologic outcome after TTM. There was a significantly better outcome in patients with pupil diameters ≤4 mm compared with those with pupils ≥4 mm.10
EEG monitoring has also been used to estimate neurologic outcome. Recent studies noted that time to continuous normal voltage (CNV) on amplitude-integrated EEG (aEEG) within 24 h after return of spontaneous circulation (ROSC) was associated with a good outcome.11,12
Researchers reported that a cutoff value of 1.17 for the gray matter to white matter attenuation ratio (GWR), comparing the putamen and corpus callosum, predicts a poor neurologic outcome with a specificity of 100% on a CT obtained immediately after ROSC (median, 69.5 min).13
Another report on GWR, averaging the basal ganglia and cerebrum, found that a cutoff value of 1.14 also predicted a poor neurologic outcome with a specificity of 100%.14
In both studies, sensitivity was low. Thus, lower normal values of GWR on brain CT obtained immediately after resuscitation can predict a poor neurologic outcome in patients with OHCA.
Other researchers reported that regional saturation of oxygen (rSO2) ≥40% on arrival at the hospital predicted good 90-day neurologic outcome with a sensitivity of 81% and specificity of 96% in patients with OHCA of presumed cardiac origin.15
Another study reported that only 2 of 446 (0.4%) patients had a good 90-day outcome if their rSO2
was low (15%) on arrival at the hospital.16
Prediction Models Consisting of Multiple Parameters
Prediction models of neurologic recovery after OHCA have been developed combining several factors. One model, consisting of 4 parameters on admission (GCS motor score, pupillary light response, spontaneous eye movements, and blood glucose), predicted a good outcome with a sensitivity of 92.0%, although the specificity was 65.0%.17
The 5-R score is another prediction model, which consists of 5 factors with a total possible score of 7: initial Rhythm, ventricular fibrillation (VF)/VT (2); starting Resuscitation; arrest to first cardiopulmonary resuscitation (CPR) attempt ≤5 min (1); ROSC, arrest to ROSC ≤30 min (2), light Reflex; pupillary light reflex in the emergency department (1); and Re-arrest, absence of re-arrest (1). A score ≥5 predicts a good neurologic outcome with a sensitivity of 82.5% and specificity of 92.3% in patients subsequently treated with TTM.18
Another prediction model uses clinical parameters that can be easily recorded at the site of CPR: age (A), arterial blood pH (B), time from starting CPR to ROSC (C), pupil diameter (D), and initial rhythm (E). The authors reported a sensitivity of 80% and specificity of 92% for predicting 6-month neurologic outcome.19
For OHCA of presumed cardiac origin, another model included variables such as age, time from collapse to ROSC, and ROSC prior to arrival at the hospital.20
Summary
Current evidence indicates which patients may not be good candidates for TTM, such as patients with a GWR ≤1.14 or rSO2
≤40. EEG monitoring has also been used to estimate neurologic outcome. However, there is no reason not to perform TTM in the majority of cases, even in patients with a GCS of M1, the presence of a dilated pupil, or a low-voltage EEG pattern. Time to CNV on aEEG within 24 h of ROSC represents a good outcome. We emphasize the use of multiple modalities or predictive models (e.g., pupil diameter, GCS, aEEG, GWR on CT, rSO2, 5-R score etc.) to evaluate brain damage and inform decision making regarding the initiation of TTM at the time of admission.
CBF and Metabolism
Jugular vein oxygen saturation (SjO2) and cerebral oxygen saturation (ScO2), such as rSO2, indicate the intracranial balance of CBF and metabolism. However, no studies have investigated tissue oxygen saturation as a guide to managing systemic circulation, respiration, or oxygen balance.
We consider ScO2
measured by near-infrared spectroscopy to represent the sum total of systemic function (e.g., airway, breathing, circulation, time course after arrest, and clinical conditions such as fever, shivering etc.). We therefore think it is important to understand the factors contributing to oxygen balance. In this section, we first explore the relationship between various cerebral oxygen parameters to offer a basis for discussion in the following section on the management of systemic parameters, and then discuss about the validity of using SjO2
or ScO2
for neuromonitoring PCAS.
Changes in CBF and Metabolism After Cardiac Arrest
Factors affecting the relationship between CBF and metabolism are shown in
Figure 2. For example, decreased blood viscosity is associated with an increase of SjO221
Blood transfusion may decline SjO2
or ScO2
by increasing viscosity, though generally it is believed that blood transfusion brings SjO2
higher. Similarly, a decrease in ScO2
may be caused by fever, shivering and insufficient sedation. Decreases in cardiac output (CO) because of cardiogenic shock, hypovolemia, low systemic vascular resistance (SVR), or with excess sedatives may also influence ScO2. In addition, hypocapnia may reduce SjO2
because of cerebral vasoconstriction. On the other hand, it may increase SjO2
because of a decline in intracranial pressure (ICP). Theoretically, we can adjust these factors by the management of seizures, hemodynamics, respiration, sedation, shivering etc.
Though these relationships are considered, cerebral pathophysiology and altered hemodynamics or metabolism makes comprehension of SjO2
or ScO2
during TTM difficult.
The pathophysiology of PCAS is focused on disturbance of calcium ion metabolism, mitochondrial permeability transition, reactive oxygen species,22
inflammation such as sepsis,23
and more. Altered intracranial blood flow after ROSC may affect ScO2. Some studies have indicated what is termed “luxury perfusion” following arrest in non-survivors.24
This is caused by CBF-CEO2
mismatch. Cerebral oxygen extraction (CEO2) is calculated by the following equation: CEO2=SaO2−SjO2/SaO2. There is a study that revealed that the cerebral metabolic rate of oxygen (CMRO2) analyzed by positron emission tomography continued to be low after ROSC and coma.25
Takasu et al revealed that SjO2
in survivors was significantly lower than that in non-survivors after ROSC.26
Bisschops et al mentioned that SjO2
increased during and after TTM, although CBF measured by transcranial Doppler did not change significantly.27
Validity of Using SjO2 or ScO2
ScO2
is oxygen saturation measured by near-infrared spectroscopy, which is an optimal, non-invasive method that measures oxyhemoglobin and deoxyhemoglobin concentrations.28
ScO2
is thought to have a similar vector to SjO2,29
instead of a difference of measurable brain area (ScO2
measures the frontal area of brain; SjO2
reflects the whole hemisphere).
In a study comparing ScO2
measured by the INVOS®
near-infrared system and SjO2
measured by jugular bulb oximetry, INVOS indicated significantly higher values at low SjO2
(≤60%) and significantly lower values at high SjO2
(≥60%) in patients after cardiac arrest.2
Another study indicated the feasibility of monitoring intracranial oxygen saturation during the post-resuscitation period.30
However, no studies have investigated tissue oxygen saturation as a guide to managing systemic circulation, respiration, or oxygen balance. In addition, the algorithm of calculating ScO2
is undocumented.28
Therefore, it is difficult to analyze ScO2, but it has the possibility of becoming a monitoring method that contributes to better neurologic outcome.
Summary
SjO2
and ScO2
may reflect systemic or cerebral conditions such as mean arterial pressure (MAP), partial pressures of oxygen and carbon dioxide in arterial blood (PaO2, PaCO2, respectively), body temperature, etc. It is difficult, however, to analyze the SjO2
or ScO2
value in patients undergoing TTM because knowing only the target value for ScO2
is inadequate without understanding other factors influencing SjO2
or ScO2. ScO2
must therefore be assessed very carefully.
Detection, Treatment, and Prevention of Seizures
Seizures are not only the result of brain injury caused by cardiac arrest but also a risk for secondary brain injury. Seizures are caused by abnormal excessive or synchronous neuronal activity in the brain.31
They are classified as generalized convulsive seizures or non-convulsive seizures. Muscle contraction and relaxation are absent in the case of non-convulsive seizures. The incidence of non-convulsive seizures after cardiac arrest is from 12% to 24% in adults and up to 47% in pediatric cardiac arrest cases. Some studies indicate that seizures occur most often within the first 8 h after ROSC according to continuous EEG (cEEG) records.32–34
Seizures are masked by neuromuscular blockage (NMB) in between 3% and 44% of cases.32,35–37
For these reasons, it is recommended to use cEEG for patients with PCAS.
Prognostication of Outcome by EEG
aEEG monitoring is a well-developed type of quantitative EEG. One study reported the possibility of predicting outcome after cardiac arrest by using aEEG. An initial low-voltage aEEG had no prognostic value, but a burst-and-suppression (BS) pattern was strongly associated with a poor neurologic outcome at any time. Electrographic status epilepticus (ESE) had 2 patterns: ESE evolving from BS had an earlier start and was resistant to treatment, whereas ESE evolving from a continuous pattern evolved later. ESE evolving from a continuous pattern reflected a less injured brain that was potentially salvageable (Figure 3).38
Recent studies suggest that the appearance of CNV within 24 h of ROSC is associated with a good outcome.11,12
Prophylactic and Therapeutic Use of Antiepileptic Drugs
Although seizures may be the result of brain injury caused by the cardiac arrest, they may themselves also cause secondary brain injury. Some investigations have revealed that antiepileptic drugs (AEDs) do not decrease the incidence of convulsive seizures or improve neurologic outcome.39,40
As there is no standard method to diagnose seizures with cEEG and the drugs may cause adverse effects, the prophylactic use of AEDs is not recommended.
There is no high-grade evidence showing a relationship between AEDs and survival or neurologic outcome.41–43
But as seizures may lead to secondary brain injury, treatment of recurrent seizures could be considered as standard therapy in comatose patients with PCAS.
Summary
The prognostic utility of aEEG has been validated. A continuous pattern or ESE with a continuous pattern is associated with good outcomes. However, there is a lack of evidence informing decisions on when or how to intervene with AEDs following the results of cEEG because of a lack of knowledge regarding the relationship between AEDs and PCAS.
Management of Systemic Parameters
Stabilization of Hemodynamics
Myocardial systolic or diastolic dysfunction44–47
and a decline in SVR46
have been observed transiently in PCAS. In this section, we discuss the management of hemodynamics after ROSC.
MAP Target Value
The question is how MAP should be managed to improve outcome in patients who have suffered a cardiac arrest. This has been investigated from various viewpoints. One study showed the best survival in patients with a MAP of 76–86 mmHg and mixed venous oxygen saturation (SVO2) of 67–72%.48
Another observational study assessed the relationship between MAP and outcome after ROSC, finding a time-weighted average MAP ≥70 mmHg was associated with a better neurologic outcome than lower levels.49
In a retrospective study, MAP ≥100 mmHg during the 2 h after ROSC was associated with better neurologic recovery at hospital discharge.50
On the other hand, a study by Young and colleagues found no relationship between higher MAP during therapeutic hypothermia and neurologically intact survival.51
Similar to early goal-directed therapy in sepsis, some investigators looked not only at MAP but also at other hemodynamic factors (e.g., central venous oxygen saturation and hemoglobin). Bundled care with goals of MAP of 80–100 mmHg, CVP ≥8 mmHg, and central venous oxygen saturation ≥65% led to better neurologic outcomes and less mortality than in historic controls.52
A similar bundle requiring MAP ≥65–70 mmHg, CVP ≥8–12 mmHg, and hemoglobin ≥9–10 g/dL showed a better survival rate to hospital discharge and neurologic outcome at 1 year.35
Those studies suggest that MAP should be kept higher than a defined threshold during the post-arrest period. Although the cerebral perfusion pressure (CPP)-CBF relationship or autoregulation of CBF is theoretically impaired,53
there is no evidence that a higher MAP causes increased ICP and worsening of outcome. Therefore, it is reasonable to aim for MAP ≥65 mmHg at all times during TTM.
How to Achieve Hemodynamic Goals
One study reported that the amount of fluid required to maintain MAP after arrest was 3.5–6 L.45,46
However, the relationship between fluid or blood products and outcome after ROSC is unclear, in contrast to what is known about albumin in sepsis.54
If fluid resuscitation alone is ineffective, it is reasonable to use vasoactive drugs.46,55
There is, however, some concern about higher MAP achieved with vasoactive agents and poor outcomes.56
The importance of heart rate control, another factor regulating CO, is controversial. In hypothermia, bradycardia occurs normally and is associated with reduced systolic dysfunction in animal models.57
Acute coronary syndrome is a common cause of cardiac arrest. In patients with ST-segment elevation or left bundle branch block on initial ECG after ROSC, the prevalence of an acute coronary lesion is more than 80%.58
Many observational studies indicate that percutaneous coronary intervention (PCI) improves survival or neurologic outcome.59,60
Some observational studies have also demonstrated better survival and functional outcome after ROSC achieved with a combination of TTM and PCI for ST-elevation myocardial infarction (STEMI).61,62
In one randomized controlled trial and pooled analysis without cardiac arrest, this combination reduced the size of the ischemic lesion of cardiac muscle in a case of hypothermia having been achieved when the coronary artery was reperfused.63,64
Studies of survival or neurologic outcome of emergency coronary intervention for non-STEMI are inconsistent, with some investigators finding it not helpful,65
but others reaching the opposite conclusion.59,60,62,66
By the way, cardiologists can participate in CPR or in intensive care after ROSC within their speciality. One study revealed that intra-aortic balloon pumping (IABP) with PCI contributed to improved neurologic outcome under cardiogenic shock after ROSC.67
Another study revealed that cardiopulmonary bypass (CPB) in addition to PCI improved neurologic outcome in patients who failed to respond to conventional CPR if the collapse-to-bypass interval was within 55.5 min.68
It may be considered to institute IABP or CPB in a patient in whom it is suspected that the cause of cardiac arrest is reversible. And cardiogenic shock should not be a reason to avoid TTM.
Summary
Recent studies indicate MAP should be maintained higher than a defined threshold during the post-arrest period. Although there is no evidence that a higher MAP causes harm, many studies indicate the MAP should be maintained above a certain threshold, the value of which has yet to be definitively determined. Therefore, we aim for MAP ≥65 mmHg at all times during TTM. To achieve this, we suggest the use of fluids, vasoactive agents, and immediate coronary interventions, particularly in cases of STEMI.
Respiratory Support
In this section, we discuss oxygenation and ventilation management after ROSC.
Oxygenation
Hyperoxemia after ROSC promotes the formation of reactive oxygen species, which can induce secondary injury in brain tissue already damaged by cardiac arrest. Both observational studies69,70
and meta-analyses71–73
show that hyperoxemia is associated with poor survival and neurologic outcome in PCAS. Although the conclusions of other studies have differed,74,75
it may be necessary to avoid hyperoxemia after ROSC.
The etiology of hypoxemia in patients with PCAS includes lung contusion induced by chest compressions, atelectasis, ventilator-associated lung injury, and others. It is no wonder that hypoxemia in PCAS may induce secondary brain damage beyond that during the arrest itself. Some studies have indicated that hypoxemia after ROSC is associated with worse outcome than normoxemia.69,74
Both hypoxemia and hyperoxemia may need to be avoided after ROSC.
Positive end-expiratory pressure (PEEP) is another factor associated with oxygenation. Interestingly, according to an observational study, protective mechanical ventilation with a lower tidal volume and higher PEEP is more commonly used after cardiac arrest. This appears to reduce the incidence of pulmonary complications, although other organs are still at risk.76
A consensus on PEEP settings for patients with PCAS is lacking, although increasing PEEP may elevate ICP.77
It may therefore be rational to maintain PEEP as low as possible as long as higher concentrations of oxygen can be avoided.
Ventilation
CO2
reactivity of the cerebral vasculature after ROSC is preserved during mild therapeutic hypothermia.78,79
Therefore, CO2
should be controlled during TTM. Hypocapnia following hyperventilation causes cerebral vasoconstriction and, based on some observational studies, it can certainly cause and/or worsen cerebral ischemia, worsen outcome, and cause injury to other organs in PCAS.80–83
On the other hand, increased PaCO2
may cause further worsening of an elevated ICP by increasing the CBF. However, evidence of the effect of hypercapnia on outcome after ROSC is conflicting.75,81–83
Although one study found that S100b concentrations decreased over time in patients with PaCO2
maintained at 50–55 mmHg but not in those with a PaCO2
of 35–45 mmHg, hospital mortality did not differ significantly between the 2 groups.84
Considering these studies, the risk of poor outcome appears to differ for hypocapnia and hypercapnia, even if differences in PaCO2
from normal are comparable. Therefore, clinicians should monitor for hypocapnia and avoid this outcome more than hypercapnia.
pH-Stat and α-Stat
Ventilation strategies during therapeutic hypothermia are based on blood gas analysis using either the pH-stat or α-stat method. In the latter, mechanical ventilation is set to achieve a physiological PaCO2
level measured at 37℃, unadjusted to the patient’s actual temperature, whereas in the pH-stat method, ventilation is set to achieve a physiological PaCO2
measured at the patient’s actual temperature. Compared with pH-stat, the α-stat has less CO2
(Figure 4).
With the α-stat method, autoregulation of the cerebral vasculature may be preserved and the incidence of cerebral microemboli can be decreased. On the other hand, this method may promote intracellular acidosis and impair cerebral oxygenation.85
One study showed that after ROSC, SjO2
was significantly lower with the α-stat method compared with pH-stat method, whereas the arteriojugular oxygen content difference and CEO2
were significantly higher, but no differences were found in CBF measured by Doppler.86
For patients undergoing TTM after cardiac arrest, it remains unclear which method is best, and either method is probably acceptable, in contrast to recommendations for hypothermic arrest for cardiac surgery.87
However, we believe hypocapnia should be strictly avoided, measured by α-stat, which is more sensitive to hypocapnia than pH-stat.
Summary
Both hypoxemia and hyperoxemia may need to be avoided after ROSC. PEEP should be kept as low as possible as long as higher oxygen concentrations can be avoided. Many studies indicate there is a risk of poor outcomes associated with hypocapnia; however, results regarding the association between hypercapnia and outcome have been conflicting. Therefore, clinicians should monitor for hypocapnia and avoid this outcome more than hypercapnia, especially when using α-stat.
Sedation Management
Sedation is widely used during critical illness. Light or minimal sedation is associated with shorter duration of mechanical ventilation, a shorter stay in the intensive care unit (ICU), and a reduced incidence of delirium.88
However, it is important to recognize that patients with PCAS with status epilepticus and uncontrolled intracranial hypertension were generally excluded from randomized trials of sedation.89
In this section, we discuss sedation for patients with PCAS.
Benefits and Harms of Sedation During TTM
Sedation may reduce secondary cerebral ischemia and decrease elevated ICP by reducing the CMRO2, CBF and cerebral blood volume.90
Sedation and analgesia also help control shivering and seizures, which is required for the induction and maintenance of TTM to reduce the risk of brain damage caused by seizures.
On the other hand, sedation makes it difficult to perform an accurate neurologic examination and clinical assessment. After arrest, residual sedation or paralysis confounds the clinical examination.91
One study showed that patients undergoing TTM (≤33℃) recovered consciousness in a mean of 3.8 days, with approximately 20% awakening after 5 days post-arrest.92
Sedative Drugs
We should consider all the potential advantages and disadvantages of sedative drugs. Propofol, currently used in many ICUs, has a rapid onset and short duration of action. It is associated with a greater risk of hypotension with cerebral hypoperfusion than other drugs, as well as the risk of propofol infusion syndrome. Although intravenous sedative agents other than ketamine decrease the CMRO2
and CBF, propofol significantly decreases CBV by causing vasoconstriction.93
Midazolam results in less hemodynamic instability than propofol, but it prolongs the duration of mechanical ventilation and length of ICU stay,94
and it may prolong the time to awakening and reduce the accuracy of the clinical examination. One study of traumatic brain injury showed that propofol and midazolam are equally efficacious in improving ICP and CPP.94
Dexmedetomidine is short-acting, provides mild to moderate sedation and analgesic effects, allows clinical assessment, and may be neuroprotective.95
Dexmedetomidine, however, frequently causes hypotension and bradycardia.
Volatile anesthetics have been reported as efficacious,96,97
but there is little data allowing comparison with other sedatives. NMB is commonly administered during TTM, resulting in more rapid achievement and maintenance of target temperature and control of shivering.98
Some studies suggest that continuous NMB has a beneficial effect and improve outcomes,99,100
but it masks seizures. EEG monitoring should be considered in comatose patients after cardiac arrest, particularly if NMB is used.8
We present the drugs commonly used during TTM in
Table 1, and sedation and anti-shivering methods in
Figure 5.
Table 1.
Drugs Commonly Used During Target Temperature Management After Cardiac Arrest
108,111,114
| Drug |
Advantage |
Disadvantage |
| For sedation and control of shivering |
| Fentanyl 25–150 mcg/h |
Short-acting |
Nausea, constipation |
| Dexmedetomidine 0.2–0.7 mcg/kg/h |
Short-acting, mild to moderate sedation
and analgesic effect, reduce delirium |
Bradycardia, hypotension |
| Benzodiazepine midazolam 2–5 mg IV, 1–10 mg/h |
Less hemodynamic instability than
propofol |
Prolonged sedation, delirium |
| Propofol 20–150 mg IV, 25–150 mcg/kg/min |
Short-acting |
Hypotension, propofol infusion syndrome |
Neuromuscular blockage: vecuronium, rocuronium,
cisatracrium |
Most effective |
Need sedation, intubation, and EEG
monitoring |
| For control of shivering |
| Acetaminophen 650–1,000 mg q4–6 h |
No sedative effect |
Liver dysfunction |
| NSAIDs |
No sedative effect |
Renal dysfunction, recent peptic
ulcer or gastrointestinal bleeding |
| Magnesium bolus 2–4 g, 12–16 g/24 h |
No sedative effect |
Renal dysfunction, check serum
levels (3–4 mg/dL) |
| Buspirone 20–30 mg q8 h |
Minimal sedative effect |
Liver or renal dysfunction, nausea,
dizziness |
| Meperidine 10–100 mg IV q4–6 h |
Effective anti-shivering |
Sedative effect, seizures, dizziness |
Monitoring the Depth of Sedation
The best choice of sedative for PCAS and the depth or dose of sedation is unknown. One review shows that moderate sedation may be effective and safe.101
In critical care, conventional sedation scoring tools such as the Richmond Agitation Sedation Scale and the Sedation-Agitation Scale are widely used and may be reasonable tools in some brain-injured patients.102
However, in patients who are comatose, deeply sedated, or who have muscle flaccidity, these tools cannot be used to measure the depth of sedation.
cEEG, quantitative EEG and bispectral index have been studied to monitor depth of sedation. Bispectral index values significantly correlated with Richmond Agitation Sedation Scale and the Sedation-Agitation Scale in brain-injured patients103
and may be useful for monitoring deeply sedated patients in the ICU.104
The EEG suppression ratio may also help assess the depth of sedation.105
However, these monitoring techniques are of questionable use in patients with PCAS because of changes in electrical activity in the brain cortex and the influence of shivering. Therefore, there is no reliable tool to monitor the depth of sedation during TTM.
Weaning of Sedation
Early interruption of sedation during TTM causes shivering. In brain-injured patients, interruption of sedation causes increased ICP.106
Therefore, a wake-up test should be avoided during the first 24 h after ROSC.107
It is suggested that the tapering of sedative infusions should not exceed 25% per day.108
In the patient at risk of brain edema or who has an elevated ICP, uncontrolled status epilepticus, and ongoing hypothermia, sedatives should not be abruptly discontinued. When weaning of sedation commences, attention should be paid to these risks with appropriate use of such tools as CT, EEG, and ICP monitoring.
Summary
Sedation is essential for patients with PCAS during TTM because of the protective effect on the injured brain. As mild sedation is not always beneficial in PCAS, we suggest sufficient sedation while paying attention to seizures, shivering, and complications. The best choice of sedative drugs is unknown, as is the best monitoring method, particularly if NMB is used. However, the amount of sedation should be tapered slowly.
Shivering Control
Shivering is a physiologic homeostatic mechanism to maintain body temperature; it is usually initiated at approximately 36℃.108
Sustained shivering causes an increased metabolic rate and CO, increasing brain oxygen consumption and ICP,109
and it increases the stress response. Shivering commonly occurs during TTM and may lead to failure to achieve or maintain adequate hypothermia. Therefore, the management of shivering, including its evaluation and treatment, is important during TTM.
Assessment and Suppression of Shivering
Shivering is often assessed with a subjective, simple, and reliable clinical scale such as the bedside shivering assessment scale (BSAS).109
Indirect calorimetry and electromyography have also been used.110
At initiation of hypothermia, skin counterwarming with non-pharmacologic methods should be considered even when surface cooling methods are used for TTM.111–113
This involves the warming of non-cooled areas of the skin (i.e., the face, hands, feet) with a warm-air blanket. Pharmacologic methods include acetaminophen, buspirone, magnesium sulfate, meperidine, fentanyl, dexmedetomidine, propofol, midazolam, and NMB. To suppress shivering and prevent prolongation of sedation and paralysis, a stepwise anti-shivering protocol during TTM is suggested.114
In general, patients with PCAS who are comatose during TTM need tracheal intubation and therefore require sedation and analgesia. To suppress shivering, a combination of methods should be used (Figure 5). The advantages and disadvantages of these drugs (Table 1) should be carefully considered, and shivering should be aggressively controlled.
On the other hand, one recent study reported that patients with the most severe brain injury have less shivering.115
Other studies have shown that the faster the target temperature is reached, the higher the mortality and the worse the neurologic outcome.116,117
It is hypothesized that patients with more severe or irreversible neurologic damage are less reactive to low temperatures, so there is less shivering117
and less requirement for NMB.118
The relationship between shivering and outcome has yet to be fully elucidated.
Summary
As shivering is a cause of failure to achieve or maintain TTM, shivering should be monitored using methods such as BSAS and aggressively controlled, including pharmacological or non-pharmacological methods.
Glycemic Control in PCAS
Glycemic control is important in the management of critically ill patients. Current concepts of glycemic control are to avoid hypoglycemia and minimize glycemic variability (GV). In this section, we discuss the characteristics of glycemic behavior after cardiac arrest and compare management of glucose in PCAS with that in other critically ill patients.
Glycemic Behavior After Cardiac Arrest
Stress-induced hyperglycemia is a well-recognized phenomenon.119
Hyperglycemia has been observed in the early phase after cardiac arrest.120
During TTM, the incidence of hypoglycemia reportedly increased significantly during the rewarming phase compared with during maintenance.121
One analysis has reported that insulin sensitivity is significantly lower and more variable during the cooling period, but it increases with warming.122
The association between early hyperglycemia after arrest and unfavorable neurologic outcome or death has been described in several studies.32,123,124
Delayed correction of hyperglycemia at admission was associated with unfavorable neurologic outcome and death.125
On the other hand, blood glucose levels at 12 h after ROSC had a non-linear association with a favorable neurologic outcome.126
Hyperglycemia after arrest may be a target for intervention, but it may merely reflect disease severity.
The effect of GV has been assessed,127
and an association between GV and mortality has been reported.128
One database study in France showed a smaller magnitude of GV was observed in patients with a good neurologic outcome compared with those with a poor outcome.129
Other studies report that increased GV was associated with increased mortality and unfavorable neurologic outcome.121,130
This suggests that attention should be paid to GV.
Target Glucose Range
An optimal glucose target range is an important issue for patients with PCAS. One study showed that mortality in patients resuscitated from VF significantly increased with increasing mean blood glucose during the first 72 h of treatment.131
Another study examined a bundle of care that included a target glucose range (90–144 mg/dL). Their protocol improved survival and neurologic outcome at hospital discharge, but how much of this effect was related to glycemic control is unknown.35
A small randomized controlled trial comparing strict (target of 72–108 mg/dL) with moderate (target of 108–144 mg/dL) glucose control revealed no benefit of strict control on mortality at 30 days, and the trial was stopped.132
Based on these studies, it appears that strict glucose control may not be necessary. However, it remains unclear what range of glucose levels is safe.
One analysis by the National Registry of Cardiopulmonary Resuscitation in the USA of patients with IHCA found interesting data suggesting that maximal glucose values above the range of 111–240 mg/dL and minimum values below the range of 71–170 mg/dL decreased survival odds in patients without diabetes.120
For patient with diabetes, a similar result was found for minimum glucose values, but they appeared to be more tolerant of higher maximal glucose values.
Summary
In patients with PCAS, the optimal target range remains unknown. However, insulin sensitivity increases and blood glucose levels decrease as body temperature rises during TTM. Further, patients with diabetes may be tolerant of higher glucose values. Accordingly, blood glucose levels should be checked frequently to avoid hypoglycemia and hyperglycemia.
